Hi again, If you are certain about your NCS operators (well refined operators) and about your envelopes, then why don't you forget about solvent flipping gamma perturbation and phase combination? Just use the experimental phases once (to compute the initial electron density maps) and throw them away afterwards? Only use the phase information (and fom's) from averaging for the next cycle. This approach has made wonders for us again and again.
Fred. > Message du 16/10/09 17:34 > De : "Pete Meyer" > A : CCP4BB@JISCMAIL.AC.UK > Copie à : > Objet : Re: [ccp4bb] DMMULTI question > > > > In the first cycle DMMULTI calculates solvent masks for the two crystals. > > Does it continously improve these masks or stick to the initial one ? > > If it acts like DM, then it updates it (using default options). But on > the other hand, DMMULTI doesn't appear to to support SOLMASK UPDATE > keywords. Kevin can speak authoritatively about this, or you could > compare SOLIN and SOLOUT and see if they're different. > > > Since DMMULTI does not apply gamma-perturbation (solvent flipping) should I > > use it to the highest common resolution (5A) only and continue with > > DM/PARROT to 4A ? > > Your mileage may vary, but a few years back I compared solvent > flattening and gamma corrected solvent flipping in PHASES on some low > resolution data and got roughly comparable results (stock solvent > flattening was slightly better for this data, though). > > > > > Thank you for your time and advice, > > > > Peter > > > > > > > > > > > >
